Rotman Research Institute, Baycrest, Toronto, Ontario, Canada.
Department of Computational Neuroscience, University Medical Center Hamburg-Eppendorf, Hamburg University, Hamburg, Germany.
Neuroimage. 2019 May 1;191:81-92. doi: 10.1016/j.neuroimage.2019.02.018. Epub 2019 Feb 7.
Reconstructing the anatomical pathways of the brain to study the human connectome has become an important endeavour for understanding brain function and dynamics. Reconstruction of the cortico-cortical connectivity matrix in vivo often relies on noninvasive diffusion-weighted imaging (DWI) techniques but the extent to which they can accurately represent the topological characteristics of structural connectomes remains unknown. We addressed this question by constructing connectomes using DWI data collected from macaque monkeys in vivo and with data from published invasive tracer studies. We found the strength of fiber tracts was well estimated from DWI and topological properties like degree and modularity were captured by tractography-based connectomes. Rich-club/core-periphery type architecture could also be detected but the classification of hubs using betweenness centrality, participation coefficient and core-periphery identification techniques was inaccurate. Our findings indicate that certain aspects of cortical topology can be faithfully represented in noninvasively-obtained connectomes while other network analytic measures warrant cautionary interpretations.
重建大脑的解剖通路以研究人类连接组已成为理解大脑功能和动态的重要努力。在体皮质间连接矩阵的重建通常依赖于非侵入性扩散加权成像（DWI）技术，但它们在多大程度上能准确代表结构连接组的拓扑特征仍不清楚。我们通过构建使用活体猕猴 DWI 数据和已发表的侵入性示踪研究数据构建的连接组来解决这个问题。我们发现，DWI 可以很好地估计纤维束的强度，而基于轨迹的连接组可以捕获诸如度和模块性等拓扑特性。还可以检测到丰富俱乐部/核心-边缘型结构，但使用介数中心度、参与系数和核心-边缘识别技术对枢纽进行分类是不准确的。我们的发现表明，某些皮质拓扑的方面可以在非侵入性获得的连接组中得到忠实的表示，而其他网络分析措施需要谨慎解释。
